Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    INDEX
    Explanations

    th

    np_max-act · gemini-2.0-flash

    This neuron detects ordinal-number suffix tokens (like “th” in “211th,” “206th,” etc.) used in legal document headings.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     Exhib
    -0.07
    يين
    -0.06
    oš
    -0.06
     quy
    -0.06
     Troy
    -0.06
     mít
    -0.06
    股份有限公司
    -0.06
    etě
    -0.06
    AM
    -0.06
    .LINE
    -0.06
    POSITIVE LOGITS
    DownLatch
    0.07
    生物
    0.07
    (pipe
    0.07
     необходимо
    0.07
    primitive
    0.06
    ιλ
    0.06
     kunt
    0.06
    !==
    0.06
     teste
    0.06
    हम
    0.06
    Act Density 0.001%

    No Known Activations